Meta Advantage+ Campaigns: What to Automate and What to Control

Meta has spent years moving advertisers from hand-built campaigns to automated ones. Sales, app and leads campaigns now start in an automation-first set-up, and the older Advantage+ Shopping and app campaign formats are being retired. That leaves most advertisers with a practical question: which decisions should you hand to Meta, and which should stay with you?

This guide answers it for people who run Meta Advantage+ campaigns for ecommerce stores, lead generation accounts and apps. It separates what Meta has documented from our own working advice, and it points out where the rules changed recently so that you can check your own account against them.

What Advantage+ means now

Meta's developer documentation describes an Advantage+ campaign as one that has three levers switched on: campaign-level budget, Advantage+ audience and Advantage+ placements. The status applies to the sales, app and leads objectives. If any of the three is restricted, the campaign is reported as having Advantage+ disabled. The status is a read-only result of how you configure the campaign, and there is no separate button for it.

In a February 2026 developer post, Meta also stated that the old Advantage+ Shopping Campaign and Advantage+ App Campaign formats can no longer be created or updated through the Marketing API. The restriction applies to every API version from 19 May 2026, and the post said remaining legacy campaigns were paused with version 26.0, September 2026. Version 26.0 was released on 29 July 2026 and its changelog does not mention that pause, so we could not confirm from Meta's documentation whether it has happened. Check the current position in your own Ads Manager before relying on it.

One consequence matters for agencies. Legacy campaigns that used the existing customer budget cap cannot carry that setting into the new structure. Meta's documentation suggests building two ad sets instead: one aimed at a custom audience of existing customers, and a second that excludes that audience.

The three levers and what each one changes

Budget

Budget is set at campaign level, and Meta distributes spend across the ad sets. The bid strategies listed in the documentation for Advantage+ status are lowest cost without a cap, cost cap, bid cap and minimum ROAS. For most accounts the practical decision is whether to start with lowest cost to gather data, or to set a cost cap or ROAS floor when you already know the margin you need.

Audience

Advantage+ audience treats your targeting as a starting point. Meta says it gives priority to people who match your suggestions, such as custom audiences, lookalikes, interests and age ranges, and then looks more widely. You can still set firm limits. According to Meta, minimum age, location, language and custom audience exclusions are controls the system will not expand beyond. Meta's leads page adds that other targeting restrictions can switch Advantage+ off.

Placements

Advantage+ placements means no placement targeting or exclusions at ad set level, so Meta can deliver wherever your ads are eligible. Account-level placement exclusions are still permitted without losing the status, which is the supported way to keep ads away from a placement you consider unsafe for your brand.

What to automate

Our recommendation, based on how Meta describes the system, is to automate the decisions that depend on signals you cannot see in real time.

Audience expansion. Let Advantage+ audience find buyers beyond your lookalikes, and treat your own lists as suggestions. Meta reports lower cost per acquisition for campaigns using it, with median improvements from 7.2% to 14.8% depending on funnel stage, each stated at 99.9% confidence.Placement choice. Unless a placement is a brand safety problem, leave delivery open. Restricting placements usually shrinks the pool the system can learn from.
Budget distribution between ad sets, once you have enough conversion volume for the system to compare them.
Creative variations, if your brand guidelines allow them. Review what the tools generate before the ads run, and turn off any enhancement that alters logos, product shots or claims.

What to keep in your hands

The inputs you give Meta decide what it optimises towards. These are the areas where we suggest you keep control.

Conversion event and value. Choose the event that reflects revenue or a qualified lead, and avoid optimising for page views or raw form submissions. For lead generation, Meta recommends connecting CRM data through the Conversions API so the system can learn which leads were worth having.
Exclusions that protect the business. Exclude existing customers or recent purchasers when you want new-customer acquisition, and exclude audiences that conflict with your terms or licensing.
Geographic and age limits that reflect where you can sell, ship or legally advertise.
Budget and bid strategy, because these set the price you can afford to pay.
Creative strategy. The system chooses between the ads you supply, so the range of offers, angles and formats is still your job.
Measurement. Agree on attribution windows and a source of truth outside Ads Manager before judging results.

How the advice changes by objective

Sales campaigns: the catalogue now works across all sales campaigns according to Meta, so a shop can run product ads inside the same structure. Keep the optimisation event tied to purchases, give the system a value to optimise towards if you sell items at very different prices, and use the two-ad-set method described earlier when you want to control spend on existing customers.

Leads campaigns: quality matters more than volume, because a cheap lead that never answers the phone teaches the system the wrong lesson. Feed back qualified-lead status through the Conversions API where you can, and keep the form short enough to complete on a phone but long enough to filter out poor enquiries.

App campaigns: the same three levers apply, but your signal depends on events reported by your measurement partner. Check that those events arrive reliably before raising spend, since the system can only optimise towards what it is told.

Limitations and risks

Automation has costs, and several are practical. You give up some of the manual levers, since Meta's documentation says certain controls change the campaign status. Moving a legacy campaign into the new structure puts it back into the learning phase according to Meta, and the existing customer budget cap has no direct equivalent. The favourable performance figures come from Meta. Third-party blogs describe many more Advantage+ layers and rules than Meta's own pages do, and some of those descriptions disagree with each other, so rely on Meta's help pages and developer documentation when you need an exact rule.

If you cannot give the system enough conversion volume, automation has little to learn from. Accounts with only a handful of purchases a month often do better with simpler structures and a higher-volume conversion event, such as add to cart, until sales data builds up. This is our working advice and not a Meta rule.

A set-up checklist

Confirm the objective (sales, leads or app) and check that your pixel or Conversions API events fire correctly and are deduplicated.
Pick the optimisation event that best represents value, and set a budget you can sustain for at least a full learning period.
Choose a bid strategy. Start with lowest cost if you have no cost target, or use a cost cap or minimum ROAS if you do.
Add audience suggestions from your first-party data, then apply only the hard limits you need.
Leave placements open and apply any brand safety exclusions at account level.
Upload a varied set of creatives with different angles, then review any generated variations before launch.
Record the launch date, the settings and the baseline numbers so you can interpret results later.

How to test it fairly

A comparison against your manual campaigns only means something when both receive comparable budgets, run over the same weeks and optimise for the same event. Run the test long enough to clear the learning phase, avoid editing the campaigns mid-test, and compare cost per qualified lead or revenue per pound spent rather than clicks. Ads Manager also offers experiment tools for split tests, and Meta's documentation describes how to set them up.

Our wider approach to automation rests on the same principle: AI handles repetitive work, and people keep judgement over strategy, budget and quality.

Frequently asked questions

Can I switch Advantage+ off completely?

You can restrict individual levers, such as placements or detailed targeting, and the campaign then reports Advantage+ as disabled. Meta's developer documentation says the full range of ad tools remains available for manual campaigns, but sales, app and leads campaigns now default to the automation-first set-up. Check what your own Ads Manager offers, because the interface changes.

Will Advantage+ work with a small budget?

It can, but results depend on conversion volume as well as budget. Meta has not published a minimum budget that guarantees success, so treat any specific figure in third-party articles as a guideline.

What happens to my old Advantage+ Shopping campaigns?

Meta has stopped creation and updates through the API and said remaining legacy campaigns would be paused from version 26.0. Duplicate or migrate them into the new structure and check the budget, audience and placement settings after the move.

Automation is now the default for sales, app and leads campaigns on Meta. The conversion goal, the data you feed in, the limits you set and the creative you supply are the parts that remain yours to manage.